Suicide
Topics Addressed
- Calculation of Conditional Probabilities
- Psychology
Suicide has been the subject of increasing study in recent
years, as psychologists struggle to understand the reasons that people
choose to take their lives. Two questions in such studies concern the
gender of the person committing suicide and the method chosen for doing so.
- Do women commit suicide at a higher rate than men, or is it the
other way around?
- Are there differences by gender in the method used?
Let's see.
Here is a 2 by 4 contingency table that categorizes U.S. suicides in 1983
by gender and method ("hanging" also includes strangulation and
suffocation). For instance, of the 28,295 suicides in the U.S. that year,
13,959 were by men using guns.
Method
---------------------------------------------------------------
Guns Poison Hanging Other | Total
===============================================================
Male || 13,959 3,148 3,222 1,457 | 21,786
Female || 2,641 2,469 709 690 | 6,509
===============================================================
Total || 16,600 5,617 3,931 2,147 | 28,295
-
What is the probability that a person chosen at random from among
these 28,295 suicide victims is male? How about female?
-
Given that the numbers of men and women in the country are about equal, what
does this say about the suicide rate among men as compared with women? Give
an explicit numerical answer (for example, "Women are more than 5 times
likelier than men to commit suicide," if that's correct).
Explain in a sentence or two what these probabilities
say about the differences, if any, between men and women in the chosen method
of suicide.
Work out the following conditional probabilities for a person chosen at
random from among these 28,295 people. Some are already computed.
- P( guns given male ) = ...
- P( guns given female ) = 40.6%
- P( poison given male ) = 14.4%
- P( poison given female ) = ...
- P( hanging given male ) = 14.8%
- P( hanging given female ) = ...
- P( other given male ) = ...
- P( other given female ) = 10.6%
George Michailides
gmichail@stat.ucla.edu